The Daily Background

(Just for clarification, this was written & published before the new Des Moines Register poll came out and wasn’t meant to specifically bash it, just to shed caution on polls in general.)

Since I’m getting asked about polling again, I thought I might as well revisit two items that I would highly recommend reading and keeping in mind:

• A note on polling (don’t believe it!) Part I (TDB)
• Polling on the Dark Side of the Moon (Pollster.com)

In short, every media organization is trying to make news with a poll that shows that things in Iowa and New Hampshire are radically different than they were last night, even though all the good pollsters out there know that:

1. The Iowa caucus is notoriously difficult to poll due to second choices and low turnout (only about 6% of the state bothers to show up– the question is which 6% will show up?)
2. Many younger, more well-educated & affluent voters aren’t home for the holidays (meaning that Clinton’s low information voters are over-represented in polls)
3. The race in Iowa is still fluid and one of the candidates can easily slip up, or the weather could be bad on January 3rd
4. NH polls are useless right now (since everything in New Hampshire will change after the Iowa caucus)

Chris Bowers has been doing some very high quality “Nomination at a glance” posts these days, and I want to highlight three important questions he raises today, regardless of whatever the polls say:

1. Iowa? Clinton holds a very slight edge in first-place choices, but that is entirely due to her advantage in ARG and Zogby, the two least respected polls among pollsters. Edwards has a favorable trend and a seeming advantage among second-place choices. Obama’s trend isn’t favorable right now, but there is also evidence to suggest that polls are currently undercounting his support in Iowa. In other words, no one has a clear advantage right now, and the top three could still finish anywhere in the top three.

2. Iowa’s impact on New Hampshire? It is hard for me to imagine a scenario where, if Clinton or Obama wins Iowa, that person does not also win New Hampshire. When it comes to Edwards, with an average Iowa bounce of 11.3% in New Hampshire, and current New Hampshrie deficits of 11.6% and 12.2%, it has become impossible to tell whether or not an Iowa win will translate into a New Hampshire win, too. Right now, Obama must wish that New Hampshire, with its enormous secular electorate, was the first state in the nation, instead of Iowa. If, for some reason, Iowa has no impact on New Hampshire, right now I would favor Obama in the state. as the trendline there is extremely favorable to him.

3. Iowa and New Hampshire’s impact on the national campaign? According to fladem, the average national swing for a sweep of Iowa and New Hampshire is 33%. This means that if either Clinton or Obama win Iowa, and thus New Hampshire, the nomination almost certainly breaks their way. If Edwards sweeps the two states, then it looks like a close two-way campaign between Edwards and Clinton. If Edwards wins Iowa, and Clinton hangs on to take New Hampshire, then it looks like Clinton wins the nomination. If Edwards wins Iowa, and Obama hangs on to win New Hampshire, then all three should have a good shot and it is anyone’s guess as to what happens next. I have to admit, the political junkie in me is kind of pulling for that result. This is great political theater, and I don’t want it to all end on Thursday night, or even next Tuesday night.

Filed under: Z-TDB Blog